A Web-Based Card Game Simulation for Optimal Strategy Development

A web-based human-computer interaction experiment simulates a game where points are scored on a board by matching pairs, opposites, or sequences of cards. The original game that inspired this experiment was modified to eliminate random card drawing. Having all cards available to each player in each move shifts the focus to skill instead of luck. An algorithm ranking all possible moves a single step in advance by points scored is pitted against human players. A modified version of the strategy favoring moves that set up future card sequences is also tested. Assigning a greater weight to playing cards that could form a sequence, even if the plan may be interrupted by the other player, allows for a more sophisticated approach while avoiding computationally intensive long-term simulation. Games are played with human or algorithm players taking the first turn to mitigate second-player advantage. Board widths of three, four, and five were tested, and only cases with a height less than or equal to the width were included to limit the scope of the study. In most cases, the algorithm focusing on chains wins against users, outperforming the baseline single-move simulation, and providing a guideline for further development.